Introduction

The steadily advancing specialization in the surgical fields, combined with a constantly increasing public awareness toward surgical competence, raises the bar for professional education. As in any surgical field, this is also true in orthopaedic trauma surgery. Educating trauma surgeons is complex, time consuming and requires financial resources, as besides lectures and theoretical items, also many manual and practical tasks need to be taught to the future surgeons [1]. Yet, not only future surgeons need education, but also surgeons who already finished their training should be introduced into newly developing techniques and implants [2,3,4]. However, as many modern medical systems find themselves under distinct economic pressure, surgical teaching during daily clinical work may be difficult to be performed on a level, necessary to fulfil the aforementioned complex tasks. To enhance surgical training, practical courses have been conducted for years. In musculoskeletal training courses, the surgeons are being educated at hands of either artificial bone samples that are closely matched to the human anatomy, or teaching is being conducted at hands of formalin fixed or fresh frozen human cadaveric specimens. With such items and specimens, the surgeons can appreciate the anatomy and perform, learn at first hand or refine surgical techniques and procedures. Yet, these methods do not allow realistic training conditions, especially in the case of orthopaedic trauma courses, fracture treatment shall be taught. Specimens with life-like fracture patterns would allow realistic training conditions, comparable with a flight simulator in aviation. Hence, the aim of the present study was to investigate a method to induce realistic fractures into human cadaveric specimens of the distal radius, aiming to make realistic fracture patterns available for such surgical training courses. The advantage would be clearly a more realistic course environment, and as published recently by our group, a possibly enhanced learning experience [5]. As it is the most commonly fractured bone, ORIF of the distal radius is one of the most typical surgical procedures in trauma surgery [6, 7]. Standard treatment of dorsally displaced fractures is volar locking plate fixation [8, 9]. The fracture itself represents a prearthrotic factor, making exact anatomical reduction and fixation an absolute necessity [10,11,12]. When facing multi-fragmentary intraarticular fractures, adaption of the numerous well-described techniques can be essential to guarantee the best possible outcome [13]. Hence, in the present paper, we aimed to reproduce realistic complex articular fracture patterns of the distal radius, by inducing a high energy axial impulse onto specifically fixed fresh frozen human cadaveric forearm specimens. As the specimens are to be used for cadaveric surgical training courses, it is mandatory that the outer soft tissue envelope remains intact, or offers typical accompanying lesions and hence, was a fundamental prerequisite for the study.

Materials and methods

Specimens

For the present study, 12 fresh-frozen human cadaveric upper extremities were available. The specimens were obtained from body donors; written consent was guaranteed. Institutional ethics committee approval was given prior to this study (VT 19-442).

The donors had an average age of 70 years (range 41–91 years) at the time of death, 3 were male and 9 were female. 5 right and 7 left specimens were available. Before testing, X-ray imaging of the specimens was used to check for marked degenerative changes or sequelae from previous fractures or the presence of implants or other foreign bodies that might influence fracture simulation.

Test-bench and fracture simulation

After thawing at room temperature, the specimens were dissected free from soft tissues over a span of 5 cm, beginning 12 cm proximal to the carpus. Then the forearm was osteotomized just proximal to the dissected area using an oscillating saw. Proceeding like this allowed to keep enough bone stock distal to the elbow joint, to be able to bring the rest of the specimens to reasonable use, which the study group sees as a prerequisite for scientific work with human cadaveric specimens. Moreover, reducing the length proximal to the carpus helps with concentrating the induced forces to the area of the desired fracture.

After the specimens were prepared in that manner, the dissected span of the radius and the ulna were potted into custom-made steel cylinders using polymethyl methacrylate (PMMA). The cylinders were equipped with a 4 mm thick baseplate, to be able to fix the construct into the fracture unit later. During the potting process, it was made sure to align the forearm bones rectangular with the baseplate of the potting cylinder, to allow precise axial force induction.

To generate the necessary forces, a custom drop-test-bench had been constructed (Fig. 1). A similar system has been used in a previous work of the study group [14]. The test-bench mainly consists of a steel frame, mounted on a 15 mm thick steel baseplate, measuring 1.5 × 1.5 m. This allows stable positioning of the test-bench and uniform distribution of the forces into the ground. The frame of the test-bench encloses a height-adjustable crossbeam and an impact beam. Within the crossbeam, two slide holes allow passing of the two stems to an impact stamp, that features two contact plates, one above and one below the crossbeam. The baseplate of the potting cylinder can be mounted to the distal contact plate of the impact stamp. In that manner, the distal end of the specimen with the hand and wrist comes to rest on the baseplate of the test-bench (Fig. 1). By levelling the adjustable crossbeam, the amount of overhang of the impacting stamp above the crossbeam can be set. When the impacting beam is dropped onto the upper contact plate of the impacting-stamp, the stamp is pushed downwards and compresses the specimen according to the amount of overhang of the stamp above the crossbeam. The compression of the specimen shall then lead to the required fracture (Fig. 2). With a calculated impact speed of 4 m/s, we are able to hit the specimen with up to 210 Joule (J) of kinetic energy \(\left( {E = \frac{1}{2}mv^{2};\,E = {\text{energy}},\; m = {\text{mass}}, \;v = {\text{velocity}}} \right)\) [14]. Damping mechanisms whatsoever, exceeding the damping effects of the specimens itself, are not needed or implemented.

Fig. 1
figure 1

Drop-test-bench setup with fixed specimen in hyperextension. a Overview, b zoomed in fixed specimen, # impact beam, + height adjustable crossbeam, *baseplate, °impact stamp, ^specimen on baseplate, < ground plate with two fixation bars

Full size image
Fig. 2
figure 2

Online supplement; series of pictures a–c showing the fracture production process from a lateral point of view; forearm is fixed on a baseplate in hyperextension with kinetic energy coming from above onto the specimen; the effect on the specimen can be observed by the scale

Full size image

To achieve fractures of the distal radius, the specimens were fixed in extension within the drop-test-bench. Therefore, the potted forearm is fixed against the inferior baseplate of the stamp, and the wrist, which is left free is positioned in extension against the baseplate of the frame. As the crossbeam is height-adjustable, different lengths of specimens can be brought into the desired position. We placed the wrist in 80° of dorsal extension. Following the fracture process, X-ray and CT imaging was conducted to analyse the injury.

Our Institutions Picture Archiving and Communication System (IMPAX) was further used to calculate Hounsfield Units (HU) as a reference to bone mineral density in our specimens [15,16,17]. According to Wagner et al. measurements were performed on three sequential coronal computed tomography scan slices at the level of the distal radial–ulnar joint [18]. The region of interest was measured via a circle to best fit the ulna head without involving the cortical surface. Mean of three measurements was then calculated for each specimen [19].

All fractures were classified independently by two trauma surgeons (KW, AH) according to the AO/ASIF recommendations [20]. Further subclassification was performed according to the recently introduced four-corner concept by Rikli et al. as a useful characterization to enable a tailored approach to the treatment [21].

Statistical analysis

Correlation analysis to age and HU, as well as kinetic energy was conducted, and two-sided t tests were performed to detect any statistically significant differences. We used descriptive statistics to summarize the means and standard deviations. The level of significance was defined as a p value of < 0.05.

Results

12 specimens were available for detailed analysis. Mean HU was 177.89 ± 72.86. All fractures were classified as type C fractures according to the AO/ASIF. There was an independent 100% agreement between the two surgeons classifying the fractures. We divided them further into subgroups of similar fracture patterns and described them additionally by the four-corner concept [21]. Each fractured specimen was suitable for osteosynthetic teaching and was thus successfully utilized in our courses.

Mean weight in our test setup was 16.3 ± 1.8 kg, mean overhang of the stamp was 0.9 ± 0.18 cm and mean kinetic energy was 134.7 ± 14.4 J. All data is summarized for each specimen in Table 1.

Table 1 Individual characteristics of each specimen
Full size table

Correlation

No statistically significant correlation could be detected between HU and required kinetic energy to achieve fracture (t value 2.088, p value 0.063), and also between HU to age (t value − 0.875, p value 0.402).

Type 23C3.2 fractures

11 out of 12 fractures (92%) were classified as type 23C3.2 fractures. Fracture patterns consisted of at least three intraarticular fracture lines, comminuted metaphysis and intact diaphysis. Carpal subluxation was witnessed in 3 cases, two dorsal subluxations and one volar subluxation. One fracture pattern consisted of a deep central impression of the articular surface, which is seldom if ever seen. A typical extension fracture pattern can be seen in Fig. 3.

Fig. 3
figure 3

AO-type C3.2 with a dorsal key corner. Fracture pattern in a fluoroscopy (a.p. and lat.), b sagittal plane, c coronal plane and d axial plane

Full size image

Each of those 11 fractures could be further classified as type C according to Rikli, meaning complete intraarticular fracture patterns [21]. A key corner could be described in 6 cases, one dorsal key corner and 5 volar. In 6 cases no key corner could be distinguished. A concomitant ulnar styloid (US) fracture was observed in 4 cases. All cases showed at least one fragment involving the sigmoid notch.

The fractures were produced with a kinetic energy between 120 and 168 J.

Type 23C3.3 fracture

One fracture showed that in addition to multi-fragmentary articular and metaphyseal portion one bony streamer to the distal diaphysis was also present. This streamer is part of a bony shell that has its origin at the dorsal metaphysis and runs towards the volar diaphysis where it ends approximately at 45 mm from the radial joint surface. No severe dislocation of diaphysis is detected. In this fracture a volar key corner can be described, as the fracture line extends coronally from the distal dorsal corner of the radius, involving Lister’s tubercle, to the volar metaphyseal portion with mainly dorsal comminution and carpal subluxation to the volar side (type C). This fracture was produced with a kinetic energy of 152 J.

Discussion

The present study shows, that by inducing a high energy impulse onto human cadaveric specimens of the forearm fixed in the hyperextension of the wrist, realistic fracture patterns can be achieved. Although, the procedure is technically demanding and a high-priced test-bench must be constructed, a clear advantage is that such “pre-fractured” specimens can be used for educational courses in trauma surgery training.

If we look at a typical surgical residency, we will come to notice that the first surgical experience as an operating doctor is often the reduction and fixation of a distal radius fracture, as they are very common and eventually are not highly demanding to fix, depending on the fracture pattern [6, 22]. Therefore, training with this kind of fracture model can facilitate early surgical education. However, the fracture patterns of our study consist of more demanding fracture types (AO type C3), enabling even more experienced surgeons to improve their technical skills, underlining the value of the presented simulation mechanism.

Only one fracture model in our study was found to be rather unrealistic, as a severe central impaction of the articular surface was found. Hyperextension and compression are the major force components that cause fractures in the wrist. Both force components can be transmitted onto the radiocarpal joint during a fall on the hand [23, 24]. An important aspect of the acting forces is the amount of extension of the wrist, as it determines the ratio between tension and compression forces on the radiocarpal joint [24, 25]. Thus, a different fracture pattern occurs by varying the degree of wrist extension [24]. Our specimens were fixed at a standardized angle of 80° wrist extension and axial load induction occurred. The outcome was 12 comparable’ but individually different fracture patterns. Bone density, individual anatomical shape and amount of axial load have a high impact on the definitive fracture pattern, comparable to real patients suffering from a distal radius fracture. Regarding this fact, every pre-fractured specimen suitable for surgical training presents a challenging and unique surgical experience. All 12 pre-fractured specimens were successfully included in our course system. Fracture reduction was, as mentioned before, demanding in many cases. The involvement of the sigmoid notch in each of our cases, demanded a precise fixation of these fragments in order to anatomically reconstruct the distal radioulnar joint (DRUJ) and achieve stability.

Further evaluation of the simulation technique of our test setup can lead to knowledge on how to simulate distal radius fractures of different severity. So far, our test setup is able to simulate reproducible type C fractures, according to the AO/ASIF. By recruitment of further specimens with manipulation of axial load, overhang of the stamp and drop height, we will be able to produce a variety of fracture patterns that can be utilized in surgical courses of different educational levels.

Moreover, our fracture setting mechanism has been validated in other parts of the body, for instance proximal and distal humerus, proximal forearm, pelvis, cephalus, metacarpal and ankle joint. Expanding on this knowledge has enabled us to install a course system of excellent surgical teaching on many different and realistic bony injuries of the human body. These simulated injuries come along with typical accompanying soft tissue lesions, as we found by dissecting the specimens in the courses. This fact even rises the educational effect of this modification of conventional cadaver courses [5].

As new implants and techniques are being introduced to the market frequently, also experienced surgeons can profit from this course system to improve their technical skills in managing new implants and learn techniques at a low-risk, but realistic surrounding.

Correlation analysis of our specimens regarding HU, age and kinetic energy showed no significant results. CT imaging was only performed after simulation to detect the exact fracture pattern. It seems unreasonable and cost-intensive to scan specimens before fracture in order to calculate HU and scan for pathologies. Conventional fluoroscopic scans prior to fracture setting showed to be sufficient in each of our cases. This step is however crucial to detect for any pathologies, that might jeopardize fracture simulation. Additionally, our specimen cohort was rather small to detect any significant correlations. Investigation of a larger cohort might show significant correlation between HU, as a parameter of bone mineral density, and required kinetic energy to produce a certain fracture [17]. So far, our fracture simulation is based on profound experience from past projects. Future investigation into the simulation of distal radius fractures will lead to increasing knowledge on exact kinetic energies. Nevertheless, this study demonstrates the first study to generate course specimens with high impact energy induction.

One of our main goals was to produce fractures with intact skin envelope in order to teach and learn surgical approaches, operational reduction and fixation technique from skin incision to suture, providing the most realistic surgical experience as possible [26, 27]. Cadaveric models remain to be the benchmark in surgical training before operating on living patients [28, 29]. As technical improvements are being introduced frequently to the market, as mentioned before, and sub-specialisation becomes more and more necessary to guarantee an up to date therapy, every surgeon needs to stay on top of his educational level. Surgical courses on cadaveric specimens, therefore represent the ideal course system to provide a significant amount of realistic teaching [30, 31].

Major drawbacks of the use of cadaveric specimens in courses, pre-fractured or not, are the costs, limits on repeated use, difficulties in procurement of specimens and high demands regarding hygienic surroundings [26, 32]. Evaluation of our courses with pre-fractured specimens emphasizes, however, that these disadvantages dissolve against a profound learning experience in realistic trauma models [5]. These findings have also been reported in a different course that provided pre-fractured cadaveric specimens [33]. Nevertheless, further investigation into the value of course systems providing cadaveric specimens is necessary [34].

In Conclusion, realistic fracture patterns of the distal radius can be produced by high energy impulse onto the forearm of cadaveric specimens, fixed in hyperextension of the wrist. All 12 specimens included in this study could be successfully implemented in surgical courses. These pre-fractured specimens can, therefore, facilitate surgical education by providing realistic fracture models with intact soft tissues. Learning to reduce and to fix a fracture on cadaveric pre-fractured specimens is able to set the benchmark in surgical trauma education, before operating on living patients. Disadvantages, such as high costs and the challenging technical procedure to simulate a fracture, dissolve against the value of providing a realistic surgical learning experience.